Alberta study: cattle stayed inside Nofence virtual fences more than 99% of the time and remembered them a year later

November 2025 — Beef cattle wearing Nofence GPS collars stayed inside virtual boundaries more than 99% of the time over two grazing seasons, a University of Alberta-led study found, and they remembered the signals a year later. Here's what happened, what it cost, and what went wrong.

Cost and outcome (as of Nov. 6, 2025)

Outcome: more than 99% containment; heifers learned the boundaries in 5–7 days. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))
Shocks: the ratio of electric pulses to audio cues fell from 17.9% in training to 5.2% while grazing. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))
Cost: collar prices weren't given; Bork calls them relatively high. (University of Alberta Folio, Nov. 6, 2025)
Herd health: pregnancy rates matched or beat Alberta's 86% benchmark. (University of Alberta Folio, Nov. 6, 2025)

What the study did

The researchers used Nofence virtual-fencing collars to run summer rotational grazing for two years. The cattle also wore leg-mounted activity sensors. In year one the animals were yearling heifers that had never worn the collars; in year two the same animals came back as first-calf cows with calves. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))

The collars track location and play an audio cue as an animal approaches a programmed boundary, then deliver a mild electrical pulse if it keeps going. They're controlled from a phone app. (University of Alberta Folio, Nov. 6, 2025)

Results

Heifers adapted to the boundaries in five to seven days. Electric pulses ran at 17.9% of audio cues during training and 5.2% during grazing. A year later, experienced cows had ratios of 1.6% during re-training and 2.2% while grazing. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))

Learning varied by age group and between individual animals. The study found no link between the number of stimuli and animal characteristics or performance. More active animals, by step count, got more audio cues. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))

In year two, cows turned away from the boundary after an audio cue 49 times out of 50, the university said. When an uncollared calf wandered outside and bawled, its mother would go to the fence and bellow but wouldn't cross. (University of Alberta Folio, Nov. 6, 2025)

Why it matters

"Virtual fencing allows the rancher to control where, when and how often the animals graze, which gives tremendous flexibility to improve pasture and forage management," said co-author Edward Bork, a professor of rangeland ecology and management. (University of Alberta Folio, Nov. 6, 2025)

Bork said the technology could sharply cut the need for costly fencing and the labour of moving herds. "In Alberta alone, there are 1.55 million breeding animals," he said. (University of Alberta Folio, Nov. 6, 2025)

"Concentrating grazing in one area for short periods of time gives a longer recovery period for soil and plants," Bork said of rotational grazing. (University of Alberta Folio, Nov. 6, 2025)

Who did it

Co-authors include Carolyn Fitzsimmons of Agriculture and Agri-Food Canada, U of A researchers Gleise Medeiros da Silva, Alexandra Harland, Francisco Novais and Maria Londono-Mendez, and John Church of Thompson Rivers University. Funders were Alberta Beef Producers, the Sustainable Canadian Agricultural Partnership, Alberta Innovates and Results Driven Agricultural Research. (University of Alberta Folio, Nov. 6, 2025) The authors declared no competing interests. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))

What went wrong

Expensive collars. Bork said virtual fencing still comes with relatively high collar costs, and that it depends on adequate cellular coverage in remote areas. (University of Alberta Folio, Nov. 6, 2025)
Individual variation. Some animals tested the boundaries much more than others. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract), University of Alberta Folio, Nov. 6, 2025)
Shocks still happen. Even trained cows got electric pulses at 1.6–2.2% of the rate of audio cues. (Harland et al., Journal of Environmental Management, May 2025 (PubMed abstract))
What we don't know yet: what the system costs per head, and how results hold with larger herds or patchy coverage.

Our follow-up question for University of Alberta

What did the collars cost per animal, and how many animals were in each cohort?

Sources

Harland A. et al., “Evaluating virtual fencing as a tool to manage beef cattle for rotational grazing across multiple years,” Journal of Environmental Management, May 2025, doi:10.1016/j.jenvman.2025.125166. pubmed.ncbi.nlm.nih.gov
Bev Betkowski, “No fences: research shows high-tech collars keep cattle from straying,” University of Alberta Folio, Nov. 6, 2025. ualberta.ca

Disclosure: [not in copy] Editor placeholder